Picture for Zheng Liu

Zheng Liu

VISTA: Visualized Text Embedding For Universal Multi-Modal Retrieval

Add code
Jun 06, 2024
Viaarxiv icon

MLVU: A Comprehensive Benchmark for Multi-Task Long Video Understanding

Add code
Jun 06, 2024
Viaarxiv icon

Compressing Lengthy Context With UltraGist

Add code
May 26, 2024
Viaarxiv icon

Are Long-LLMs A Necessity For Long-Context Tasks?

Add code
May 24, 2024
Viaarxiv icon

MS MARCO Web Search: a Large-scale Information-rich Web Dataset with Millions of Real Click Labels

Add code
May 13, 2024
Viaarxiv icon

Extending Llama-3's Context Ten-Fold Overnight

Add code
Apr 30, 2024
Viaarxiv icon

Understanding Privacy Risks of Embeddings Induced by Large Language Models

Add code
Apr 25, 2024
Figure 1 for Understanding Privacy Risks of Embeddings Induced by Large Language Models
Figure 2 for Understanding Privacy Risks of Embeddings Induced by Large Language Models
Figure 3 for Understanding Privacy Risks of Embeddings Induced by Large Language Models
Figure 4 for Understanding Privacy Risks of Embeddings Induced by Large Language Models
Viaarxiv icon

ChatRetriever: Adapting Large Language Models for Generalized and Robust Conversational Dense Retrieval

Add code
Apr 21, 2024
Viaarxiv icon

Metacognitive Retrieval-Augmented Large Language Models

Add code
Feb 18, 2024
Figure 1 for Metacognitive Retrieval-Augmented Large Language Models
Figure 2 for Metacognitive Retrieval-Augmented Large Language Models
Figure 3 for Metacognitive Retrieval-Augmented Large Language Models
Figure 4 for Metacognitive Retrieval-Augmented Large Language Models
Viaarxiv icon

BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models

Add code
Feb 18, 2024
Figure 1 for BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models
Figure 2 for BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models
Figure 3 for BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models
Figure 4 for BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models
Viaarxiv icon